Abstract. In this paper we introduce a new error measure, integrated reconstruction error (IRE) and show that the minimization of IRE leads to principal eigenvectors (without rotat...
A fundamental problem in neural network research, as well as in many other disciplines, is finding a suitable representation of multivariate data, i.e. random vectors. For reasons...
Abstract. Nonlinear component analysis is a popular nonlinear feature extraction method. It generally uses eigen-decomposition technique to extract the principal components. But th...
We propose a simple, isometry invariant pattern matching algorithm for an effective data reduction useful in layout-related data processing of very complex IC designs. The repeata...
Mariusz Niewczas, Wojciech Maly, Andrzej J. Strojw...
Walsh Transforms measure all sources of nonlinear interactions for functions that have a bit representation. There can be exponentially many nonlinear interactions and exactly com...
Soraya B. Rana, Robert B. Heckendorn, L. Darrell W...